Skip to main content
Internet Archive's 25th Anniversary Logo

Full text of "NVO and the LSB Universe"

See other formats




NVO Workshop 

Caltech, June 13-16, 2000 

NVO and the LSB Universe 

James M. Schombert 

Department of Physics, University of Oregon, Eugene 


O . Abstract. 

^^ I There is tremendous scientific potential in a National Virtual Ob- 

^ . servatory, particularly for projects that need to mine large databases for 

Mh| rare or unusual objects. However, the NVO will also make an impact on 

any project, large or small, the requires a mixture of datasets to explore 
a wide range of astrophysical phenomenon. In this article I discuss the 
^^ ■ influence of the NVO on research into the formation and evolution of low 

surface brightness (LSB) galaxies. In particular, I present the prelimi- 
^ ■ nary results from an NVO-style project that combines the DPOSS and 

O ! 2MASS datasets to search for giant disk galaxies. 



O ■ 1. The Dataset Revolution 


Astrophysical problems seem to increase in complexity with each successive gen- 
Qh' eration. Observationally, new wavelengths and new flux limits bring about new 

Q . phenomenon that demands more telescope time and begs theoretical interpreta- 

iH ' tion. In the last 15 years we have seen an explosion in the amount, wavelength 

^ . coverage and diversity of our datasets that have lead to numerous discoveries, 

but have also buried us in the sheer quantity of information. 

Our community has also grown parallel to our data growth, but most of the 
high powered observational tools still lie in the possession of a few institutions. 
This disparity in big telescope resources has been offset, in a large part, by 
5^ \ the formation of national data centers and the distribution of analysis software. 

Now an astronomer, regardless of the size of their home institution, can have 
access to high quality data and produce cutting edge science. With the addition 
of small grant programs (i.e. NASA's ADP program), an astronomical industry 
developed during the last two decades and discovery has moved from the hands 
of the few to the hands of the many. One only need to compare the impact 
of HST science on the astronomical community to that of Keck to see how the 
existence of non-proprietary datasets can push forward science. 

As datasets have grow larger, there has been a strong emphasis on data 
mining and computational power. However, intelligent and cleverly designed 
projects depend more on access to direct tools rather then sophisticated al- 
gorithms. Thus, many astronomical projects today involve teams of scientists 
who bring together the diverse talents needed to attack the details of massive 
amounts of data. This is the goal of NVO, to provide the infrastructure and 
intellectual support for astronomical programs that compliment our already ex- 
isting observational and theoretical foundations. 


* • • • 

• ■* 




» ■>'■■•! 



Pu^p" ' •' 


Figure 1. A subset of DPOSS normal (left panel) and smooth with a 
5 pixel gaussian filter (right panel). A LSB disk galaxy is barely visible 
in the normal frame. A second LSB dwarf galaxy is discovered in the 
upper right-hand section of the smoothed frame. Both objects where 
detected in HI at Arecibo, the LSB disk has a velocity of 5,550 km/s, 
the dwarf is at 2,500 km/s. 

Within the NVO concept there lies the goal of removing the division be- 
tween the type of science and the type of data. Indeed, the greatest benefit the 
NVO may bring to us is to eliminate the specialization that divides our fields of 
research (stellar versus extragalactic) and between the wavelength regions (di- 
vided mostly due to the technology used to observe within the spectral regions 
of interest). As an example, one area of astronomy that could derive enormous 
benefit from the NVO is research into low surface brightness (LSB) galaxies. 

2. LSB Universe 

One area that is particularly challenging to the observational world is the uni- 
verse of LSB objects. We can only study what we know to exist, and with respect 
to galaxies that means the object must reside in some catalog. While there has 
always been a push to find the faintest objects (meaning the lowest in mass as 
stellar luminosity maps into baryonic mass) or the most distant (meaning the 
closest to the galaxy formation epoch), it has only been recently that there has 
been much concern for objects with low luminosity density. 

The LSB realm is interesting to many astrophysical problems. For exam- 
ple, star formation is normally a phenomenon associated with high gas density 
environments. Yet, LSB galaxies display many characteristics that indicates a 
history of recent star formation. Thus, studies into their past will probe star 
formation in new parameter space. The range of galaxy properties requires an 
examination of the LSB universe because galaxies at the extreme ends of the 
mass spectrum (dwarf and giant) tend to be LSB in nature. LSB galaxies also 

differ from their brighter cousins in that there gas masses often exceed their 
stehar masses. Thus, baryon counts of the Universe at high redshift wih be 
underestimated without some knowledge of the distribution of LSB galaxies. 

In some sense, the lack of pursuit of LSB research is technical in nature. To 
find the faintest, or most distant, galaxies becomes a simple process of building 
larger and larger collecting surfaces. However, achieving fainter levels of surface 
brightness to explore the LSB universe battles against the natural glow of the 
night sky and is not overcome with larger pieces of glass. Space imaging has the 
immediate advantage of getting above the sky glow, but the emphasis in space 
has been on small and faint, so pixel sizes to take advantage of high resolution 
images work against LSB objects by reducing the number of counts per pixel. 
Even in the non-optical portions of the spectrum the emphasis is always first on 
the detection threshold of point sources rather than design for sky brightness 
(see, for example, the 2MASS survey). 

The first expeditions into the LSB universe where taken, of course, by 
Zwicky who hypothesized on the existence of 'hidden' galaxies as a counter to 
Hubble's notion that the galaxy luminosity function is gaussian in shape. By the 
1960's, numerous galaxy catalogs by Arp, Sandage, van den Bergh (DDO) and 
de Vaucouleurs (RC2) had defined the Hubble classification system. These cat- 
alogs were primarily defined by HSB galaxies, but there were always appendices 
or notes concerning 'diffuse' objects, usually assumed to be nearby dwarfs. 

Disney (1976) was the first to place the concept of galaxy visibility in an 
analytic form and to demonstrate that the mean central surface brightness of 
our galaxy catalogs was, in fact, a function of the natural sky brightness and not 
imposed by astrophysics. While the importance of this work is recognized today, 
galaxy evolution was a relatively new field at the time and the study of LSB 
galaxies remained in the background (no pun intended) until the mid-1980's. 

3. LSB Detection 

There was very little that the typical observational astronomer could do about 
the night sky problem until the mid-1980's with the evident of the Second Palo- 
mar Sky Survey. While the sky brightness had only degraded since the first 
Sky Survey, the finer emulsions and deeper plate material allowed for, at least, 
a cursory examination of the difference between the angular limited UGC and 
the new plate material. This resulted in the PSS-H LSB catalogs (Schombert & 
Bothun 1988, Schombert et al. 1992, Schombert et al. 1997) which were visual 
surveys, but demonstrated an increase of one mag arcsec"^ to the old catalogs. 
More importantly, it provided a jumpstart to the LSB field by simply providing 
a new list of objects in which to study. 

The first visual surveys were extremely crude but spurred a more exacting 
search for LSB's using CCDs (O'Neil, Bothun & Cornell 1997, Dalcanton et al. 
1997). Flattening is always a key parameter for LSB galaxy detection. Most 
CCDs on 1-meter class telescopes are sky limited in a few minutes of exposure 
time. The critical component in finding LSB galaxies, and measuring fluxes, is 
how well you know the sky value and how fiat (on large and small scales) you can 
make your data. Transit CCD surveys offer the best method for sky flattening, 
since they allow the sky to pass through each pixel which is then summed for 

V = 9500 km/s v = 1 1800 km^ 

W50 = 500 km/s , W50 = 400 km/s 

J65 1_13_12_270_76 1 J65 1_07_02_28 1048 


V = 1 1000 km/s . V = 1 1700 km/s 

W50 = 425 km/s W50 =450 km/s 

4 . . " ^ 

J584 17 19 874 074 J580_15_19_746_088 


Figure 2. Four AGN Malin objects discovered from a combined 
search of DPOSS and 2MASS datasets. The near-IR 2MASS cata- 
logs provide the low luminosity AGN sources (due to their anomalous 
colors) typical to the Malin class. The optical catalog is fast filtered 
for the presence of an LSB disk. 

the entire scan. Long scans can suffer from temporal variations, but these can 
at least be quantified. 

LSB detection is functionally a very difficult problem. The standard pro- 
cedure is to detect all the HSB objects, mask them out, smooth and run the 
detection scheme again. Unfortunately, this will usually remove any LSB galax- 
ies with bright bulges. They will be in the original catalog, but their LSB nature 
may not be known from whatever parameters are stored in the catalog. 

One example of a smooth and search algorithm is shown in Figure 1. Here a 
sample of the sky from a J plate of DPOSS is shown in its raw form. A single LSB 
disk galaxy is obvious near the center of the frame. Smoothing with a gaussian 
filter enhances the detectable of the LSB disk and also discovers a second LSB 
galaxy in the top right portion of the scan. Both objects were detected in H I 
at Arecibo with velocities of 5,550 km/s and 2,500 km/s respectfully. 

The smoothed image demonstrates the two major difficulties in automatiz- 
ing LSB detection algorithms. The first is that the number of pixels distorted by 
stellar sources is extremely high at the surface brightness levels of interest. It is 
practically impossible to detect, in an automatic fashion, LSB objects near stars 
(although the human brain seems to carry out the task fairly well). Second, 
the uncorrelated background noise varys substantially over even this small piece 
of sky. This makes a systematic survey, to specific threshold levels, a technical 
challenge (what NASA would call completeness and reliability). 

4. AGN Malin Search 

One of the most intriguing classes of LSB galaxies is the supergiant disk systems, 
the Malin class. The prototype to this class, F568-6, was discovered from a 
visual search of the PSS-II (Bothun et al. 1990). While low in central surface 
brightness {jXo = 23. 4i? mag arcsecs"^), F568-6 is by no means low in luminosity 
{Mb = —21.1) nor low in total or HI mass. The Malin class contains the 
largest galaxies in the Universe, yet are notorious difficult to find and catalog 
(Sprayberry, Impey & Irwin 1996). 

One of their properties provides a promising avenue for the cataloging of 
a significant number. Most Malin class galaxies have a weak AGN in their 
core. The current theory is that the copious gas supply in the disk provides the 
fuel for a central engine, even if at a low intensity (Schombert 1998). While 
the weak AGN appears as a point source, its near-IR colors would distinguish 
it from a stellar SED. Unfortunately, a survey of the sky in the near-IR will 
not, by itself, identify the Malin objects since the sky brightness at 2.2 ^um is 
2000 times higher than in the optical making their disk regions invisible. Thus, 
this project requires a 'virtual observatory', in this case the combination of two 
existing databases, DPOSS (optical) and 2MASS (near-IR). 

The procedure is straight-forward, first isolate all the objects in the 2MASS 
catalog with non-thermal colors (i.e. outside some boundary defined by normal 
stars). Second, search the near-IR source positions on the blue plates of DPOSS 
with a fast area scan. A series of circular apertures are placed around the point 
source then tested against the local sky. LSB galaxy detection is effective if 
only a particular region is being tested against the background since varying 
diameters are checked which maximize the signal from the LSB disk versus the 
signal from sky. 

A preliminary search was undertaken last winter using eight plates from 
DPOSS that contained some fraction of 2MASS coverage (about 3 square de- 
grees). Forty candidates were produced of which ten of these objects were 
searched with the new, upgraded Arecibo telescope. The Gregorian system 
at Arecibo has a much wider velocity range, a critical element since the large 
Malin objects tend to be at velocities greater than 8,000 km/s. Eight of the ten 
candidates were detected at 21-cm. 

Four of the detection's DPOSS images are shown in Figure 2. All have 
the characteristics AGN nucleus surrounded by a LSB disk. All eight also have 
H I widths in excess of 350 km/s (the typical spiral has a rotation width of 250 
km/s). Assumingly these galaxies follow the baryonic TF relation (McGaugh et 
al. 2000), then their masses will exceed 10^^ Mq. 

5. Conclusions 

It is universally recognized that the NVO would be a powerful tool for the 
astronomical community. A particular emphasis will be placed on the need for 
the NVO to make the most efficient use of the large datasets in our present 
holdings and to future projects. However, perhaps one of the most common 
uses of the NVO will be of the one human/one workstation type project. 

One such project I have described herein, the search for AGN Malin galaxies 
typifies my vision of how small projects will use the NVO. The merger of multi- 
wavelength datasets and simple tools, combined with a researchers experience 
in the astrophysical phenomenon, was used here to achieve a catalog of a new, 
and exciting, type of galaxy. 

While each of our individual research interests may reap the benefits of the 
NVO, there is no doubt that the sum of the contributions of many small projects 
will also service to build the framework of the NVO. A system that we hope will 
be wavelength and research field independent. 

Acknowledgments. I wish to thank my collaborators in the LSB universe 
(G. Bothun, J. Eder, S. McGaugh and K. O'Neil) for their insights. Many 
different telescopes have contributed to the LSB searches, but a special thanks 
is given to the Arecibo scientists and staff for their speedy upgrade which has 
increased LSB's visibility by many orders of magnitude. I also wish to thank the 
organizers of the conference for inviting me and the NASA grant (NAG5-6109) 
which has supported my LSB programs. 


Bothun, G., Schombert, J., Impey, C. & Schneider, S. 1990, ApJ, 360, 427 
Dalcanton, J., Spergel, D., Gunn, J., Schmidt, M &: Schneider, D. 1997, AJ, 

114, 635 
Disney, M. 1976, Nature, 263, 573 

O'Neil, K., Bothun, G. & Cornell, M. 1997, AJ, 113, 1212 
Schombert, J & Bothun, G. 1988, AJ95, 1389 

Schombert, J, Bothun, G., Schneider, S. & McGaugh, S. 1992 1988, AJ103, 1107 
Schombert, J., Pildis, R. k Eder, J. 1997, ApJ, 481, 157 
Schombert, J. 1998, AJ, 116, 1650 
Sprayberry, D., Impey, C. & Irwin, M. 1996, ApJ, 463, 535